OpenAI discloses on October 14 Higgsfield integration: GPT-6 Astra now supports generating AI videos longer than 10 minutes
OpenAI on October 14 published a Higgsfield customer case study showing that after integrating GPT-6 Astra, the platform now supports generating AI videos longer than 10 minutes, up from a previous one-minute cap, with the model handling directing, Blender 3D modeling and visual consistency.
OpenAI on October 14 published its latest progress with AI video generation platform Higgsfield: after integrating GPT-6 Astra, Higgsfield now supports generating AI videos longer than 10 minutes, up from a one-minute cap. GPT-6 Astra supports creative agents that call external tools and handle complex work like 3D modelling, which Higgsfield applies to the hardest part of long-form video: keeping characters, scenes and visual details consistent across shots. Higgsfield has used the workflow to generate more than 10-minute educational videos, cutting significant manual production time; the official Higgsfield tutorial showcases a roughly 100-scene, 10-minute finished piece focused on whole-piece visual consistency. Higgsfield founder and CEO Alex Mashrabov said the company's annualized revenue grew from $1 million to $1 billion over the past 18 months, and the startup closed a $400 million round in August. OpenAI's case study noted the next step is to explore voice-to-video creation with GPT-Live-1; the model itself does not generate video pixels directly but rather takes on the directing, shot planning, 3D modelling and editing steps.